Vol. 166: 237-245,1998
MARINE ECOLOGY PROGRESS SERIES
Mar Ecol Prog Ser
1
Published May 28
l
Daily larval growth and RNA and DNA content of the
NW Mediterranean anchovy Engra ulis encrasicol us
and their relations to the environment
Alberto Garcia*, Dolores Cortes, Teodoro Ramirez
Instituto Espanol de Oceanograiia. Centro Oceanografico de Malaga, Puerto Pesquero sln, E-29640 Fuengirola. Malaga, Spain
ABSTRACT: Anchovy Engraulis encrasicolus larvae, collected from 2 different sites in the Western
Mediterranean Sea (the Gulf of Lions and the Cdtalan Sea),were analyzed for daily growth rates and
RNA and DNA content. An exponential function describes the growth pattern. The larvae were shown
to have a growth rate of 0.56 mm d-' at 8 mm (standard length). Anchovy larvae from the Gulf of Lions
showed higher daily growth rates than larvae from the Catalan Sea. In accordance, sagittal size, measured in terms of otolith area and radius, of larvae from the Catalan Sea were smaller than those of
larvae from the Gulf of Lions. RNA/DNA ratio ranged from 1.09 to 7.79, and the mean RNA/DNA ratio
was 4.49. A n c h o ~ ylarvae from the Gulf of Lions were shown to have a higher slope in the RNA
(pg larva-') versus DNA (pg larva-') relationship than larvae from the Catalan Sea. This indicates that
the former were in better condition. However, due to the high variability of the RNA/DNA ratio, no significant differences in RNA/DNA ratio were found between the 2 areas. A preliminary study of the relationship between environm~nla1factors 2nd th? XN,I!DNA ratio was peifolllled ior each region. 'The
RNA/DNA ratio of larvae from the Catalan Sea correlated negatively with temperature and pos.i.tively
with relative fluorescence and organic matter, while in the Gulf of Lions there was a positive correlation with temperature and no relation with relative fluorescence or organic matter.
KEY WORDS: Anchovy - Larvae . Growth . Otolith . RNA . DNA . Condition
INTRODUCTION
Variability in recruitment is determined by survival
during the early llfe stages of fish. Starvation is a
primary cause of mortality during larval stages (May
1974, Leak & Houde 1987). Poor nutritional condition
can increase vulnerability to predation and hamper
growth rate (Buckley 1984), increasing the time of exposure in the more vulnerable length classes (Folkvord
& Hunter 1986).
The existence of a daily rate of deposition of increments on otoliths has been extensively used for age
determination and growth analysis in the early life
stages of a large number of species (Ceffen 1982, Campana & Nielson 1985, Suthers et al. 1989), and more
particularly in the highly variable stocks of the clu-
0 Inter-Research 1998
Resale of full article not permitted
peoid species (Methot 1983, Moksness & Wespestad
1989, Alvarez & Alemany 1992). Analysis of otolith
microstructure may also be applied to study the influence of ambient factors on growth of individuals. Environmental conditions, mainly temperature and food
availability, can produce differences in the increment
widths and in the distance to the hatch check in the
otolith of a species inhabiting different marine ecosystems (Campana & Neilson 1982, 1985). Increment
width differences observed in the North Sea herring
have enabled the distinction of a n autumn- from a
spring-spawning herring (Moksness & Fossum 1991).
Different growth rates can be related to the specific
productivity of different spawning grounds, as observed by Alvarez & Alemany (1992).They compared
daily growth rates of the Iberic Atlantic sardine spawning in upwelling and non-upwelling areas.
The amount of DNA, which is the carrier of genetic
information, is considered constant in somatic tissues
238
Mar Ecol Prog Ser
and is directly related to the cell numbers of an individual, and thus is a measure of size for an individual
(Folkvord et al. 1996), whereas the amount of RNA in
cells is directly proportional to the protein synthesis
rate. Therefore, the ratio RNA/DNA is a n index of the
cell's metabolic rate. The nutritional condition of fish
larvae can be determined from the RNAIDNA ratio
(Buckley 1984, Clemmesen 1988) or by the amount
of RNA in tissues (Ferguson & Danzmann 1990, Mathers et al. 1992). Larvae subject to starvation register
lower RNNDNA ratios in comparison to fed larvae.
Moreover, RNA/DNA ratios have been used to measure recent growth rate in fish larvae (Buckley 1984,
Hovenkamp 1990, Hovenkamp & Witte 1991, Westerman & Holt 1994, Folkvord et al. 1996). A decrease in
this index is accompanied by growth decrease during
periods of starvation. Experimentation under controlled conditions has shown that RNA/DNA ratios are
affected by ambient factors such as temperature
(Buckley 1982, 1984) and prey density (Buckley 1980,
Wright & Martin 1985).
Under an EU financed project on the NW Mediterranean anchovy, a fish egg and larva survey (MAD0792) was carried out from June 27 to July 25, 1992.
Within the sampled area, 2 hydrographic regions, the
Gulf of Lions and the Catalan Sea, were differentiated.
The Ligunan-Provenzal current brings colder waters
of Atlantic origin into the Gulf of Lions in a NE-SW
direction along the edge of the continental slope (Salat
& Cruzado 1981, Mlllot 1990),separating the less dense
shelf waters from the more dense waters offshore. This
current distributes the freshwater runoff of the Rh6ne
River over the shelf, thereby producing an enrichment
of its waters (Cruzado & Velasquez 1990). In addition,
there is wind-induced coastal upwelling, which further
enhances its productivity. Its influence is observed on
the northern coasts of the Catalonian region off the
Cape Creus during spring and the onset of summer,
thereby differentiating the northern region from the
rest of the Catalan Sea coasts (Maso & Duarte 1989).
This study presents the results of the analysis of NW
Mediterranean anchovy daily larval growth rates and
their condition. The particular environmental settings
of the 2 regions covered led us to analyze the differences in daily larval growth and condition between
the 2 areas. Differences in growth are based on the
size-at-age relationship and the allometric differences
observed between somatic and otolith growth (otolith
radius and area). The differences observed in the
larval condition between the 2 regions are based on
the linear relationship between RNA and DNA content. Finally, as a preliminary study the influence of
the 2 main environmental factors, namely feeding
availabrlity and temperature, on RNA/DNA ratios was
examined.
MATERIAL AND METHODS
Sampling at sea. The fish egg and larva survey MAD0792 (Fig. 1) was carried out in the NW Mediterranean
(the Gulf of Lions and the Catalan Sea), on board
the RV 'Garcia del Cid', from June 27 to July 25, 1992,
coinciding with the peak anchovy spawning season.
Plankton sampling was carried out with a Bongo net
with a 40 cm mouth opening and a 200 pm mesh.
Another Bongo net with a 12 cm mouth opening and a
53 pm mesh was mounted over it to sample the microzooplanktonic component. The average duration of
tows was 15 min. The sampling stations were dlstributed in transects perpendicular to the coast, with a distance of 10 or 12 nautical miles between them,
depending on the observed anchovy egg abundances.
The offshore limits of the transects were determined bp
the presence of anchovy eggs. At all stations, temperature, salinity and fluorescence data were measured in
CTD vertical profiles. The anchovy larvae sampled
were collected at the stations shown in Fig. 1. Larvae
were sorted as soon as they were captured. Larvae for
nucleic acid analysis were sized and frozen in liquid
nitrogen. Larvae that were destined for otolith microstructure analysis were preserved in 5 % formaldehyde
buffered with borax at pH 8-9.
Otolith mounting and age reading. Before otolith
extraction, larvae were measured (standard length,
SL) by means of a public domain Image analysis soft-
' Gulf of Lions
Barcelona
Ebro
":---
.
X
X
.
X
%
.
I
.
Catalan Sea
Fig. 1 MAD-0792 survey, location of sampling stations for
Engraulis encrasicolus eggs and larvae in NW Mediterranean
Sea. ( X ) Stations where larvae were collected for the present
study
Garcia et al.. NW Mediterranean anchov y dally larval growth rates and condition
ware system running in a Macintosh framework, developed by the U.S. National Institute of Health (NIH;
Bethesda, MD). Sagittae were extracted under a binocular stereomicroscope with polarizing filters and
mounted using nail lacquer as a mounting medium.
For the daily age readings, the sagittae were viewed
on a video monitor and daily rings were counted using
the OTO software designed by Andersen & Moksness
(1988).Area of each sagitta was measured by means of
the aforementioned image analysis system. The criteria
followed for the daily increment readings was based on
a n otolith intercalibration exercise for clupeoid larval
otoliths (Anonymous 1993). The distance from the
nucleus to the first distinct increment for this species
was established as approximately 6 pm and the increments were counted at the widest radius of the otolith.
Larval size was corrected for shrinkage d u e to fixation and manipulation by using Theilacker's (1980)
procedure.
Extraction and quantification of nucleic acids.
Nucleic acids were extracted and analyzed by utilizing
the method described by Clemmesen (1988),allowing
for some n~odifications.Anchovy larvae were homogenized at 0°C in Tris-buffer (pH = 8) containing proteinase K (0.2 mg ml-') using ultrasonic pulses (2 X 10 s)
from a Branson Sonifier 250.
The individual content of RNA and DNA was determined with a Perkin-Elmer LS-5 spectrofluonmeter,
using specific dyes for nucleic acids. Ethidium bromide
(excitation: 324 nm, emission: 589 nm) was used for
the total quantification of nucleic acids (RNA + DNA),
while bisbenzirnide (H-33528) (excitation: 352 nm,
emission: 447 nm) was used for DNA determination.
Disposable methacrylate cuvettes were employed
for the fluorescence measurements. Estimates of the
DNA and RNA content of each larva were taken from
calibration curves established using standard DNA
(calf thymus) and RNA (yeast).
Data treatment. The larval anchovy growth models
tested were based on a n exponential fit and the LairdGompertz model using the Fishparm statistical package (Saila et al. 1988). The OPS software (Moksness
1990) was used to estimate daily growth rates (mm d-l).
This program uses the files created by the OTO age
reading program.
The samples used in the present study a r e representative of each of the 2 defined regions, the Gulf of Lions
and the Catalan Sea. The border between these 2
regions is defined as Cape Creus.
In order to find out whether otolith growth and somatic
growth were significantly different between the 2
regions, a n ANCOVA (analysis of covariance) test was
applied using log otolith radius as a covariate. Recent
otolith growth was measured using the mean width of
the last 5 increments. Average width of the last 5 incre-
239
ments was plotted against the logarithm of otolith radius.
The residuals of this linear regression were used as a n
index of recent otolith growth (Hovenkamp 1990).
ANCOVAs were also used to test the differences in
log RNA by region, with log DNA and larval size as
covariates. A l-way ANOVA (analysis of variance) was
carried out to test the difference in RNA content by
size class in the 2 regions sampled.
Correlation analyses were performed to analyze
individually the effects of temperature and food availability (as measured from relative fluorescence a n d
microzooplankton abundance expressed in terms of
organic matter) on RNA/DNA ratios by region. We
consider the 20 m depth level, according to the observations of Palomera (1991) on NW Mediterranean
anchovy larval vertical distribution, as representative
in the assessment of the environmental influence of
fluorescence a n d temperature on R N N D N A ratios.
The environmental data used for these analyses are in
the EU FAR Final Report on Project MA 3.730 (Garcia
1994). All statistical tests were carried out to a significance level of 5 % (Zar 1974).
RESULTS
Anchovy daily larval growth
A total of 156 anchovy larvae were analyzed, 79 captured in the Gulf of Lions a n d 77 in the Catalan Sea.
The size frequency distribution of the larval population
analyzed varied from 4 to 22 mm.
The daily larval anchovy growth of the whole population sampled fitted a n exponential model. Nevertheless, for comparative purposes, the Laird-Gompertz
growth model proposed by Zweiffel & Lasker (1976)
for the northern anchovy Engraulis mordax was tested
as well. Both exponential a n d Laird-Gompertz models
led to simllar results. The estimated daily growth rates
for a n 8 m m larva were similar (0.56 a n d 0.57 m m d-')
for the 2 models. Length at intercept, Lo, of both curves
fits were 4.56 and 4.57 mm respectively. (Table 1).The
Table 1. Engraulis encrasicolus. Parameters of the growth
models applied: length at t = 0 (L,), specif~cgrowth rate ( A o )
and exponential decrease of growth ( a )
Exponential
Laird-Gompertz
La"d-Gompertza
Laird-Gompertzb
Lalrd-~ompertza.b
a
LO
Ao
4.56
4.57
3.40
3.76
3,40
P
P
0.0731
0,1291
0.1560
0,1868
0 0032
O 0480
0.0589
0781
' L o flxed at 3.4 mm
bData from Palomera et al. (1988)
0.868
0.876
0.857
-
240
Mar Ecol Prog Ser 166: 237-245, 1998
-
Catalan Sea
- o- - Gulf of Lions
0
2
A
b
20
40
60
80
100
Radius (pm)
Catalan Sea
- - o- - Gulf of Lions
.
- o-
-
..
Catalan Sea
Gutf of Lions
I
-
0
5
10
15
20
25
Daily increments
Fig. 2. Engraulis encrasicol~~s.
(a] Size-at-age data for larvae
of the Gulf of Lions, y = 4.339exp(O.O76x),R' = 0.901, and the
Catalan Sea, y = 5.106exp(O.O61x), R2 = 0.790. (b) Grotvth
rates estimated by OPS vs age for larvae of the Gulf of Lions
and the Catalan Sea
0
0
5000
l0000
15000
20000
Area (pm2)
Fig. 3 . Engraulis encrasicolus. Allometnc relationships by
region. (a) Otolith radius vs larval size: the Catalan Sea,
y = 0 . 5 1 7 ~ ' ~ " R2
! = 0.891; the Gulf of Lions: y = 0 . 8 8 0 ~ ' . ~ ' ~ ,
R' = 0.924. (b) Otolith area vs larval size. the Catalan Sea,
y = 0 . 6 7 5 ~ ~R2
~ =~ 0.899;
",
the Gulf of L~ons:y = 1 . 8 1 0 ~ ~ . ~ ' ~ ,
R2 = 0.926
Laird-Gompertz model has been previously applied
to Mediterranean anchovy by Palomera et al. (1988).
When analyzing the size-at-age data by region, differences between areas were observed (Fig. 2a). These
differences were also observed in daily anchovy larval
growth rates (Fig. 2b). In order to test whether these
differences were significant, an ANCOVA to test for
the effect region on log size, with larval age as covariate, was applied. The results of this analysis showed
that the interaction effect was significant, indicating
differences in the daily growth rates (Table 2) of the
larval anchovy populations from the 2 regions. Larvae
from the Gulf of Lions showed higher growth rates.
The relative growth of larval length to otolith size,
whether expressed as radius or otolith area, showed a
potential relationship (Fig. 3). An ANCOVA to test for
the effect of region on log SL (size), with log otolith
radius as covariate, was used. The interaction effect of
region X log otolith radius was significant (Table 3).
Results of this analysis showed that anchovy larvae
from the Gulf of Lions had smaller otoliths than those
of the Catalan Sea.
Table 2 Engraulis encrasicolus ANCOVAs to test for the
effect of region on log SL (size) of anchovy larvae, with age as
covariate. ' p < 0.05, "p < 0.0001; ns: not significant
Table 3. Engraulis encrasicolus. ANCOVAs to test for the
effect of reglon on log SL (size) of anchovy larvae, with log
otolith radius as covariate. "p < 0.0001; ns: not significant
P
Region
Age
Interaction
Region
Log radius
Interaction
1
1
1
0.01
12.23
0.09
0.01
12.23
0.09
1.95ns
1758.19"
13.08"
Garcia et al. NW Mediterranean anchovy daily larval growth rates and condition
In addition, average width of the last 5 increments
were linearly related to the logarithm of otolith radius
( y =-4.295 + 4.798x, p 0.0001, R2 = 0.817). The residu a l ~of this regression were significantly higher for
larvae of the Gulf of Lions (ANOVA, p < 0.05).
a
--0--
24 1
DNA
RNA
Larval condition
A total of 177 larvae were analyzed for RNA and
DNA content, of which 83 were from the Catalan Sea
and 94 from the Gulf of Lions. Larval size varied from
6 to 10 mm.
Larvae were assigned to size classes with the aim of
analyzing their individual variability in RNA a n d DNA
content within each size class. DNA and RNA increased exponentially with larval size (y = 0.001 18x3",
R2 = 0.992; y = 0 . 0 0 1 8 7 ~R*
~ ~= ~0.994,
,
respectively)
(Fig. 4a). Moreover, the 2 nucleic acids showed a
strong linear relationship with each other (y = -1.097 +
5.324x, R2 = 0.861). The RNA/DNA ratio correlated
positively with larval size (y = 2.449 + 0.267x, p <
0.0005) (Fig. 4b), although d u e to the high variability of
the index within each size class. The coefficient of the
linear regression was rather low (R2 = 0.073). Individual RNA/DNA ratios varied from a minimum of 1.09 for
a 6 mm larvae to a maximum of 7.79 for a 10 m m larva
(Table 4). The mean RNA/DNA ratio of the whole population sampled was 4.49 (SD = 1.28).
To verify differences in larval condition between
the 2 regions, ANCOVAs to test for the effect of
region on log RNA, with log DNA and size as covariates, were applied. The interaction effects of region X
log DNA and region X larval size were significant
(Table 5). These results show that the increase of
RNA is larger per unit DNA a n d per unit size for larvae from Gulf of Lions than those caught in the
Catalan Sea (Fig. 5a, b).
The l-way ANOVA used to test the differences of
RNA content by size class showed that, except for the
8 mm size class, there was a significant difference in
RNA content between the 2 regions (Table 6). However, the ANOVA did not show regional differences in
Size (mm)
Fig. 4. Engraulis encrasicolus. Relationship between (a) nucleic
acid content and larval size and (b) RNA/DNA ratio and larval
size
the RNA/DNA ratio in any of the size classes (p > 0.05),
probably d u e to its higher variability (Table 6 ) .
The relationship of RNA/DNA ratio with the surrounding environmental characteristics of each region
was analysed. The Catalan S e a is characterized by
higher temperatures than the Gulf of Lions region
(18°C a n d 16.5"C mean temperature a t 20 m were registered in this survey, respectively). RNA/DNA ratios
of the Catalan Sea anchovy larvae showed a negative
correlation with temperature ( y = 16.121 - 0.651x, p <
0.05, R2 = 0.064), while larvae from the Gulf of Lions
Table 4. Engraulis encrasicolus. RNA content (pg larva-'). DNA content (pg larva-') and RNA/DNA ratios by size class.
Mean, standard deviation, maximum and minimum values a n d number of larvae analyzed ( n ) are given
Size
(mm)
n
6
7
8
9
10
41
52
35
30
19
Mean
RNA
(SD) Max.
Min.
Mean
2.47
4.92
8.74
12.29
20.18
(1.18)
(1.75)
(2.54)
(3.61)
(5.31)
0 24
2.09
4 10
6.26
8.47
0.61
1.15
1.92
2.64
3.73
5.52
10.90
15.10
19.40
31.20
DNA
(SD) Max.
Min.
Mean
(0.24)
(0.25)
(0.36)
(0.58)
(0.56)
0.21
0.70
1.40
1.80
2.80
4-13
4.28
4.55
4.68
5.39
1.10
1.80
3.00
3.90
4.60
RNA/DNA
(SD) Max.
Min.
(1.47)
(1.27)
(1.04)
(1.07)
(1.20)
1.09
2.28
2.41
2.66
2.92
6.83
6.63
7.55
6.93
7.79
Mar Ecol Prog Ser 166: 237-245, 1998
242
Table 5. Engraulis encrasicolus. ANCOVAs to test for the
effect of region on log RNA of anchovy larvae, with log DNA
and larval size class as covariates. 'p < 0.05;"p c 0.0001;
ns: not significant
Table 6. Engrauhs encrasicolus. ANOVA of log RNA and
RNA/DNA by reglon. ' p < 0.05, "p < 0.001; ns: not significant
Size class
df
SS
MS
F-value
1
1
1
1
1
0.47
0.15
0.01
0.18
0.08
0.47
0.15
0.01
0.18
0.08
7.18'
6.63'
0.85ns
14.47
5.50'
Log RNA
Size class
Region
Size class
Interaction
1
1
1
6
7
8
9
10
0.03
16.81
0.81
Log DNA
Region
Log DNA
Interaction
1
1
1
1.84 X 10-'
19.65
0.08
1.84 X I O - ~
19.65
0.08
0.Olns
1051.25"
4.12'
show a positive correlation (y = -7.937 + 0.743x, p <
0.005, R2 = 0.148).
Individual RNA/DNA ratios of larvae from the
Catalan Sea showed a positive correlation with fluorescence ( y = 2.268 + 41.483x, p < 0.0001, R2 = 0.235)
RNMDNA
6
7
8
9
10
1
1
1
1
1
1.99
1.99
6.58 X 1 0 - ~ 6.58 X 10-'
3.95
3.95
0.12
0.12
0.45
0.45
"
0.9211s
3 90 X 1 0 - ~ n s
3.95ns
0.lOns
0.30ns
and organic matter (y = 2.741 + 0.444x, p < 0.0001,
R* = 0.228), while these trends were not observed with
anchovy larvae from the Gulf ot Lions ( p > 0.5).
DISCUSSION
Larval growth
Catalin Sea
L
log DNA(pg larva-')
0
-0.8
I
S
6
7
8
9
10
11
Size (mm)
Fig. 5. EngrauLis encrasicolus. Linear relationships by reglon
of (a) log RNA vs log DNA and (b)log RNA vs larval size class
The growth pattern of NW Mediterranean anchovy
larvae can be described by an exponential model and
also by the Laird-Gompertz growth model. Both
models have proven to be adequate in describing
the daily growth pattern within the range of larval
sizes analyzed. The exponential growth model has
been frequently used in other engraulid species, such
as Engraulis m o r d a x and Anchoa mitchilli (Kramer &
Zweiffel 1970, Cowan & Houde 1989), while the LairdGompertz growth model proposed by Zweiffel &
Lasker (1976) for the northern anchovy E. mordax
has been applied previously on NW Mediterranean
anchovy by Palomera et al. ( 1 9 8 8 ) .
Both of the models showed similar estimates of daily
growth rates at 8 mm (0.56 and 0.57 mm d-l) and L,
estimates (4.56 and 4.57 mm). These estimates differ
from the L, value observed by Palomera et al. (1988)
and Regner (1985).These authors established, an L, =
3.4 mm, through biological observation of size at yolksac absorption. Different L. values can greatly influence the Laird-Gompertz growth parameters, specific
growth rate and the exponential decrease of growth
( A oand a, respectively). When L. was fixed at 3.4 mm,
the parameters A. and a. obtained were comparable to
the results obtained by Palomera et al. (1988)(Table 1).
With regard to the estimations of daily growth rates,
Palomera et al. (1988)reported a growth rate of 0.9 mm
d-' at 8 mm at a mean temperature of 20°C, whereas
our estimates at this size is 0.56 mm d-' This difference
can be mainly attributed to a temperature dependent
Garcia et al.: NW Mediterranean anchovy daily larval growth rates and condition
factor, taking into account that temperature greatly
influences growth rates (Buckley 1984). Our survey
was carried out in June-July, during which a mean
surface temperature of 18°C was registered, whereas
larvae were collected by Palomera et al. (1988) in a
warmer period (August) and a warmer area of the
Catalan Sea, off the southwestern coasts facing the
Ebro River mouth. In comparison, Re (1987) reported
integrated anchovy daily growth rates of this same
species inhabiting estuaries of the Portuguese coasts
from 0.3 to 0.4 mm d-' at temperatures ranging from
16.5 to 19.5"C. Methot & Kramer (1979) calculated a
growth rate of 0.37 mm d-' for an 8 mm SL northern
anchovy (Engraulis mol-dax) at a mean tenlperature
of 15°C. Therefore, the warmer temperatures of the
Mediterranean Sea could account for the observed
higher daily anchovy larvae growth rates.
The ANCOVA test applied to the growth model
showed faster growth rates in larvae sampled in the
Gulf of Lions region. The relatively higher growth
rates of the anchovy larval population from the Gulf of
Lions compared with that from the CatalAn Sea may be
due to the higher productivity off the Gulf of Lions. The
latter region is characterized by several frontal systems, namely the shelf-slope front produced by the
Ligurian-Provenzal current, the haline fronts produced
by the RhGne River, which can be detected from as
far off as the southern end of the Cape Creus in the
form of freshwater plumes (Caste11611 et al. 1985), and
wind-induced coastal upwelling which produces higher
243
nutrient and phytoplankton production (Cruzado &
Velasquez 1990). The microzooplankton (expressed
as organic matter) abundance is a potential feeding
resource for the anchovy larval population. Microzooplankton abundance in the Gulf of Lions was higher in
general than in the Catalan Sea (Fig. 6).
On the other hand, deficient growth rates cause a
longer time of exposure in the larval sizes which are
more vulnerable to predation (Folkvord & Hunter
1986); hence larvae with higher growth rates from the
Gulf of Lions could have higher survival rates than
larvae from the Catalan Sea. This hypothesis is consistent with the estimated anchovy larval mortality rates
of Palomera & Lleonart (1989), who observed notably
lower mortality estimates in the northern Catalonian
regions (in the vicinity of the Cape Creus) in comparison with the southern anchovy larval population.
The existence of uncoupling between otolith growth
rates and somatic growth rates has already been reported (Secor & Dean 1989).This happens when otolith
growth occurs even when somatic growth is near zero.
As a consequence, slower growing larvae have larger
otoliths because they are capable of accumulating
more increment counts, although producing narrower
increment widths. In the present allometric analysis,
the relationship between larval size and otolith radius
and area showed significant differences between the 2
regions. For the same larval size classes, the Catalan
Sea anchovy larvae had notably larger otoliths, accounting for slower growth rates in larvae from the
Catalan Sea. In addition, analysis of the residuals
showed that larvae from the Gulf of Lions had higher
recent otolith arowth than larvae from the Catalan Sea.
>l4 mg m-3
10-14 mg m-?
Larval condition
8-1 0 mg m-3
6-8 mg m-'
4-6 mg m-'
2-4 mg
<2 mgm-'
39"
I
0°'
Ol0
,
02"
l
I
04"
I
05"
E
Fig. 6. Microzooplankton distribution, measured as organic
matter (data acquired in FAR Project ~ ~ 3 . 7 3DG
0 , XIV), in
the Gulf of Lions and the Cataliin Sea
RNA/DNA ratios have been used as an index of recent growth rate in fish larvae (Buckley 1984). Larvae
.under good nutritional conditions have higher growth
rates, RNA/DNA ratios and RNA content. It was interesting to test whether the conclusions obtained from
the otolith microstructure analysis with the larval condition analysis were in agreement. The difference observed between the slopes of log RNA versus log DNA
regressions of the 2 populations showed that the increase in RNA per unit DNA and per unit length was
higher for larvae from the Gulf of Lions. A higher rate of
increase in RNA per unit DNA indicates a higher metabolic rate, hence higher rates of protein synthesis.
Therefore, larvae from the Gulf of Lions were in better
condition than larvae from the Catalan Sea. This could
be due to the effect of different environments on larval
condition. These results are in line with the growth
differences observed from size-at-age data.
244
Mar Ecol Prog Se
The mean RNA/DNA rat10 of the whole anchovy
population sampled lvas 4 49 As a companson Clemmesen (1989) found a mean RNA/DNA index of 3 for
field-captui e d Atlantic herring There is no experimental data on RNA/DNA estimates for anchovy Engraulis
encras~colusm t h which a cntical RNA/DNA value could
be established Clemmesen (1994)established a critical
value of 1 2 for l 0 mm herrlng larvae starved for 6 to 9 d
From the present analysis, 74 6 O/o of the anchovy larvae
had RNA1DN.A ratios between 3 to 6, 12 4 % had values
over 6, a n d only 1 1 % had values under l 2 RNA/DNA
ratios m field-captured larvae are generally higher than
those observed in reared larvae (Buckley 1984 Rob~nson
& Ware 1988, Clemmesen 1989) It is normal to expect
few larvae with a n RNNDNA ration under 1 2 in the
field, since these are more llable to die
In spite of the fact that RNA/DNA ratios show high
vanability within each size class they showed a
positive relationship with slze Clemmesen (1994) also
observed a n increase of RNA/DNA ratios in well-fed
Baltic herring larvae Larger larvae have greater nutritional reserves in their tissues (Hakanson 1989) which
could explaln the increase of the index with size, making them more resistent to prolonged penods of starvation without substantially affecting their RNA/DNA
index (Clemmesen 1989) This increase in RNA/DNA
ratio can also be d u e to cellular slze enhancement
(hypertrophy) (Goss 1966) or to a n allometric Increase
of tissues with different RNA/DNA ratios (Bulow 1987)
It has been shown that the RNA/DNA ratio is related
to water temperature (Buckley 1984) The analysis of
individual estimates of RNA/DNA indices of both
larval populations indicated a posltive correlation with
the prevailing temperature for larvae from the Gulf of
Lions system ( p < 0 005) whereas it correlated negatively with temperature for larvae from the Catalan
Sea waters ( p < 0 05) Buckley (1982) found that temperature was the dominant growth factor when larvae
have adequate food supplies However, when the temperature range is narrow, inversely, food availability
becomes the predominant factor for growth a n d condition (Buckley 1979, 1984)
The relation between food availabihty and RNADNA
ratio was analyzed Results of these analyses showed
that RNA/DNA in the Catalan Sea was highly correlated xvith microzooplankton organlc matter and relative fluorescence (p < 0 0001) Food availability could
represent a limiting factor for larval condition off the
Catalan Sea coasts In contrast, food availability does
not show a relationship with RNA/IINA in larvae from
the Gulf of Lions, a n d therefore it does not seem to be
a determinant factor for larval condition in larvae from
the Gulf of Lions, as the correlation analysis ( p 0 5)
indicates These results show regional differences as
observed in larval growth
Conclusions
The growth pattern of NW Mediterranean anchovy
larvae IS well defined by a power function Larval oto11th microstructure was successfully used to differentiate the growth rates of 2 population in different environments This d~fferencewas also shown by the
relationship between somatic growth and otolith
growth For the same larval size classes, the Catalan
Sea anchovy larvae had larger otolith sizes The rate of
increase in RNA per unit DNA was higher in larvae
from the Gulf of Lions, therefore indicating better condltion The difference observed in larval growth rates
and condition between the 2 regions implies a n environmental cause Based on this hypothesis, it \v111 be
necessary to take into account envlronmental factors in
future studies on larval growth and condition in the
field
Acknowledgements The authors are ~ n d e b t e dto the Department of Analyt~calC h e m ~ s t r yand R ~ o c h e n l ~ s tof
r y the Unlvers ~ t yof Malaga We also thank Angel Carpena who helped
\ n t h much of the otohth lab work a n d Jose M a r ~ aOrtlz d e
Urblna for h ~ s t a t ~ s t ~ c a ld v ~ c e Also, thanks to Dr Isabel
Palomera for p r o v ~ d l n gthe data on m~crozooplanktonand to
Dr Anld Folkvord and Dr Erlend Moksness for t h e ~ valuable
r
comments on the manuscnpt F~nally,we thank the flnanclal
ald of the EU for supportlnq the FAR %TA-3 730 project on the
NW Mediterranean anchovy
LITERATURE CITED
Alvarez F, Alemany F (1992) Regional growth differences In
sardlne (Sardina pjlchardus Walb ) larvae from Cantab n a n a n d Gallclan coasts. ICES CM H:9
Andersen T, Moksness E (1988) Manual for reading daily
Increments by the use of a computer programme. Flodevig e n Meldlnger 4 D~rectorateof Fisheries, Flodevigen Biol o g ~ c aS
l t a t ~ o nArendal
,
Anonymous (1993) Report of the Clupeoids O t o l ~ t hMicrostructure Workshop. Inf Tec Inst Esp Oceanogr 1.45
Buckley W (1979) Relationsh~psbetween RNA-DNA ratio,
prey d e n s ~ t yand growth rate In Atlantic cod ( C a d u s
rnorhua) larvde. J Flsh Res Bd Can 36:1497-1502
Ruckley LJ (1980) Changes in ribonucle~ca c ~ d deoxyribo,
nucleic acid and protein content durlng ontogenesis In
tvlnter flounder, Pseudopleuronectes americanus, and
effect of starvation Fish Bull US 77:703-708
Buckley LJ (1982) Effects of temperature on growth a n d bloc h e m ~ c a lcornposit~onof larval winter flounder (Pseudopleuronectes amencanus). Mar Ecol Prog Ser 8 181-186
Buckley LJ (1984) RNA-DNA ratlo: a n lndex of larval flsh
growth In the sea. Mar Blol 80:291-298
Bulow FJ (1987)RNA-DNA ratios lnd~catorsof growth In flsh: a
revlew In. Summerfelt RC, Hall GE (eds)The a g e and growth
of flsh The Iowa State University Press, Ames, p 45-64
Campana SE, Neilson J D (1982) Daily growth Increments in
otol~thsof starry flounder (Platichthys stellatus) a n d the
Influence of some envlronmental vanable in t h e ~ producr
tlon C'an J Flsh Aquat Scl 39:937-942
Campana SE, Neilson JD (1985) Microstructure of fish otolith
C a n J Flsh Aquat Sci 42:1014-1032
Garcia et al.: NW Mediterranean anchovy daily larval growth rates and condition
245
Castellon A, Salat J , Maso M (1985) Some observations on
Rhone river fresh water plume In the Catalan coast.
Rapp Comm Int Mer Medit 29(3):119-120
Clemmesen Ch.1 (1988) A RNA and DNA fluorescence technique to evaluate the nutntional condition of individual
mal-ine fish larvae. Meeresforschung 32:134-143
Clemmesen CM (1989) RNA/DNA ratios of laboratory reared
and wild herring larvae determined with a highly sensitive fluorescence method J Fish Rio1 35:331-333
Clemmesen C (1994) The effect of food availability, age or
size on the RNNDNA ratio of individually measured herring larvae: laboratory calibration. Mar Biol 118:377-382
Cowan JH, Houde EH (1989) Growth, growth vanability and
survival of Bay anchovy (Anchoa rnltchilli) larvae in mesocosm enclosures. ICES CM L:3
Cruzado A, Velasquez 2 (1990) Nutrients and phytoplankton
in the Gulf of Lions, Northwestern Mediterranean. Cont
Shelf Res lO(9-11):931-942
Ferguson MM, Danzmann RG (1990) RNA/DNA ratios in
white muscle a s estimates of growth in rainbow trout held
at d~fferenttemperatures. Can J Zoo1 68:1494-1498
Folkvord A , Hunter JR (1986) Size specific vulnerability of
northern anchovy (Engraulis mordax) larvae to predation
by fishes. Fish Bull US 84:859-869
Folkvord A, Ystanes L, Johannessen A, Moksness E (1996)
RNA-DNA ratios and growth of herring (Clupea harengus)
larvae reared in mesocosms. Mar Biol 126:591-602
Garcia A (1994) Northwestern Mediterranean anchovy: distribution, biology, fisheries and biolnass estimation by different methods. Final Report. FAR Project MA3.730. DG-XIV,
Directorate General for Fisheries of the Commission of
The European Communities
Geffen AJ (1982) Otolith ring deposition In relation to growth
rate in herring (Clupea harengus) and turbot (ScophthalImus maximus) iarvae. Mar H101 71:317-326
Goss RJ (1966) Hypertrophy versus hyperplasia. Sciences
(NY) 153:1615-1620
HakGnson JL (1989) Analysis of lipid components for determining the condition of anchovy larvae, Engraulis mordax. Mar Biol 102:143-151
Hovenkamp F (1990) Growth differences in larval plalce
Pleuronectes platessa In the Southern Bight of the North
Sea a s indicated by otolith increments and RNA/DNA
ratios. Mar Ecol Prog Ser 58:205-215
Hovenkamp F, W ~ t t eJIJ (1991) Growth, otolith growth and
RNA/DNA ratlos of larval plaice Pleuronectes platessa in
the North Sca 1987 to 1989. Mar Ecol Prog Ser 70:105-116
Kramer D, Zweiffel JR (1970) Growth of anchovy larvae
(Engraulis mordax Girard) in the laboratory as influenced
by temperature CalCOFI Rep 14.84-87
Leak JC. Houde ED (1987) Cohort growth and survival of bay
anchovy Anchoa rnichilli larvae in Biscayne Bay, Florida.
Mar Ecol Prog Ser 37 109-122
Maso h4, Duarte CM (1989) The spatial and temporal structure of hydrographic and phytoplankton biomass heterogeneity along the Catalan coast (NW Mediterranean).
J Mar Res 47313-827
Mathers Eh?, Houlihan DF, Cunningham MJ (1992) Nucleic
acid concentrations and enzyme activities a s correlates of
growth rate of the saithe Pollachius virens: growth-rate
estimates of open sea fish. Mar Biol 112:363-369
May RC (1974) Larval mortality in marine fishes and critical
period concept. In: Blaxter JHS (ed) The early life history
of fish. Springer Verlag. Berlin, p 3-1 9
Methot RD Jr (1983) Seasonal variation in survlval of larval
northern anchovy, Engraulis mordax, estimated from aged
d~stributionof juveniles Fish Bull US 81:741-750
Methot RD Jr. Kramer D (1979) Growth of northern anchovy
Engraulis mordax, larvae in the sea. Fish Bull US 77:
413-423
Mlllot C (1990) The Gulf of Lions' hydrodynamics. Cont Shelf
Res 10:885-894
Moksness E (1990) Manual for OPS (Otolith Population Statistics). Fladevigen Meldinger 5. Directorate of Fisheries,
Fladevigen Biological Station, Arendal
Moksness E, Fossum P (1991) Distinguishing spring- and
autumn spawned herring larvae (Clupea harengus L.) by
otolith microstructure. ICES J h4ar Sci 48-61-66
Moksness E, Wespestad V (1989) Ageing and back-calculated
growth rates of Pacific herring, Clupea pallasi, larvae by
reading daily otolith increments. Fish Bull U S 87509-530
Palomera I (1991) Vertlcal distribution of eggs and larvae of
Engraulis encrasicolus In stratified waters of the Western
Mediterranean. Mar Biol 111:37-44
Palomera I, Lleonart J (1989) Field mortality estimates of
anchovy larvae, Engraulis encrasicolus, in the western
Mediterranean. J Fish Biol 35(Suppl A):133-138
Palomera I, Morales-Nin B, Lleonart J (1988) Larval growth
of anchovy, EngrauLis encrasicolus, in the Western Mediterranean Mar Biol 99.283-291
Re P (1987) Ecology of the planktonic phase of the anchovy
Engraulis encrasicolus (Linnaeus, 1758) wlthin the Mira
estuary. Invest Pesq 51:581-598
Regner S (1985) Ecology of planktonic stages of the anchovy,
Engraulis encrasicolus (Linnaeus, 1758), In the central
Adriatic. Acta Adriat 26:l-113
Robinson SMC, Ware DM (1988) Ontogenetic development
of growth rates in larval pacific herring, Clupea harengus
pallad, measured with RNA/DNA ratios in t h <trait
~
of
Georgia, Bntish Columbia. Can J Fish Aquat Sci 45:
1422-1429
Saila SB, Recksierk CW, Praeger MH (1988) Basic Fishery
Science Programs. A compendium of microcomputer programs and manual of operation. Elsevier Science Publishing CO,New York
Salat J , Cruzado A (1981) Masses d'eau dans la Mer. Mediterranee Occidentale: Mer Catalane et eaux adjacentes.
Rapp Comm Int Mer Medit 27:201-209
Secor DH. Dean J M (1989) Somatic growth effects on the
otolith-fish size relationship in young pond-reared stripet
bass, Adorone saxatllls. (Walbaum). Can J Fish Aquat Sci
46:113-121
Suthers IM. Frank KT, Campana SE (1989) Spatial cornparison of recent growth in postlarval Atlantic cod (Gadus
morhua) off southwestern Nova Scotia: Inferior growth in a
presumed nursery area. Can J Fish Aquat Sci 46(Suppl 1):
113-124
Theilacker G (1980) Changes in body measurements of larval
northern anchovy, Engrauljs mordax, and other fishes due
to handllng and preservation. Fish Bull US 78:685-692
Westerman M, Holt GJ (1994) RNA:DNA ratio during the
critical period and early larval growth of the red drum
Sciaenops ocellatus. Mar Biol 121:l-9
Wnght DA, Martin FD (1985) The effect of starvation on
RNA:DNA ratios and growth of larval striped bass,
Morone saxatiljs J Fish Biol 27:4?9-485
Zar J H (1974) Biostatlstical analysis. Prentice-Hall, Englewood Cliffs, NJ
Zweiffel JR, Lasker R (1976) Prehatch and posthatch growth
of fishes. A general model. Fish Bull US 74509-621
Editorial responsibility: Otto Kinne (Editor),
Oldendorf/Luhe, Germany
Submitted. August 11, 1997; Accepted: Mai-c1123, 1998
Proofs received from author(s): April 29, 1998